Early results of reverse total shoulder arthroplasty using a patient-specific baseplate to address severe glenoid deficiency

Abstract

Background

The purpose of this study was to describe outcomes of patients undergoing reverse total shoulder arthroplasty (rTSA) using a patient-specific, custom glenoid component to address severe glenoid deficiency.

Methods

Retrospective chart review identified patients at a single institution undergoing rTSA using the glenoid vault reconstruction system (VRS) between 2017 and 2022. Radiographic evaluation, range of motion and patient-reported outcome (PRO) measures, complications, and re-operations were assessed.

Results

Fourteen shoulders were included. There was 100% implant survivorship of the glenoid baseplate at mean follow-up of 26.6 months. Mean range of motion improved in forward elevation (62–106 degrees), abduction (41–100 degrees), and external rotation (11–36 degrees). In 7 of 13 patients available for PRO collection, the mean final Visual Analog Pain Scale (VAS) score was 1.29, Single Assessment Numeric Evaluation (SANE) score was 72.14, American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form (ASES) score was 77.14, and Penn Shoulder score was 72.26.

Conclusions

Use of this custom glenoid resulted in encouraging clinical and radiographic outcomes, with no failures in implant survivorship seen at early follow-up. Larger prospective studies with longer-term follow-up should be undertaken in order to better determine the efficacy and longevity of this implant.

Study Design

Retrospective case series; Level of evidence, 4.

Introduction

Management of glenoid deformity and bone loss remains a challenge in reverse total shoulder arthroplasty (rTSA) and is recognized as the most common etiology for aseptic baseplate loosening in the primary and revision setting.^1,2 This risk of early mechanical failure is frequently associated with an inability to position components with adequate inclination and version, and obtain stable baseplate fixation in the setting of glenoid deformity.^3–5 Numerous techniques exist for the management of glenoid bone loss, depending upon the respective morphology and degree of bone loss present, including eccentric reaming, the use of metallic augmented baseplates, and structural grafting.^4,6–13 Inherent limitations compromising the positioning and stability of the baseplate remain with the use of these techniques in the presence of more severe deformities, with greater degrees of bone loss, or in the revision setting where structural autograft may not be readily available, leading to mixed clinical results in this setting.^4,6–15

Recently, patient-specific, 3D-printed, custom glenoid components have been utilized to address severe glenoid deficiency in rTSA.^{4,12,13,16–22} The glenoid vault reconstruction system (VRS) with the comprehensive reverse shoulder arthroplasty system (Zimmer-Biomet, Warsaw, IN, USA) remains one of the few commercially available systems for this purpose in the United States.^{4,12,13,16–22} Currently, a relative paucity of literature exists evaluating the use of the VRS.^4,12,13,16 Reported clinical and radiographic outcomes with the use of this system have been encouraging, however, further assessment is needed to evaluate the use of the VRS in order to determine the safety, efficacy, and survivorship of this implant system.^4,12,13,16 Therefore, the purpose of this study was to describe the clinical and radiographic outcomes at short-term follow-up of a series of patients at a single academic institution who underwent rTSA using the VRS to manage severe glenoid bone deficiency.

Methods

Study design

Institutional Review Board (IRB) approval was obtained to conduct this retrospective evaluation of all patients undergoing primary and revision rTSA at a single, academic institution using the glenoid VRS with the comprehensive reverse shoulder arthroplasty system (Zimmer-Biomet, Warsaw, IN, USA) for massive glenoid bone loss (Figure 1). Patients were identified retrospectively by querying the respective CPT codes for primary and revision rTSA (23472, 23473, 23474) and subsequently reviewing the implant records of all potentially eligible patients. All rTSAs were performed by one of two fellowship-trained, high-volume shoulder arthroplasty surgeons between 2017 and 2022. The indication for consideration to use the VRS for both surgeons was the presence of combined and uncontained glenoid bone loss or glenoid dysplasia extending up to, or medial to, the base of the coracoid (Table 1). ²³ In each respective patient, the determination for the indication to use the VRS was made using the clinical judgment of the attending orthopedic surgeon and informed consent by the patient (Figure 2).

Figure 1.

Preoperative anteroposterior (a), scapular-Y (b), and axillary (c) right shoulder radiographs of a patient with end-stage shoulder arthritis and severe glenoid retroversion.

Table 1.

An overview of the demographics, indications, and outcomes of included patients in the current study.

Patients	Age	Sex	Laterality	BMI	Primary/Revision	Prior surgeries	Bone loss pattern	Indication for surgery	Implant survivorship	Follow-up (months)	Notes
1	68	F	Right	32.9	Revision	1	Global	Glenoid wear status post prior hemiarthroplasty	Yes	29
2	83	F	Right	38.3	Primary	0	Central and posterior	Rotator cuff tear arthropathy	Yes	29.3
3	79	F	Right	30.2	Revision	1	Anterior and central	Aseptic loosening of prior anatomic TSA	Yes	29.6
4	60	F	Right	29.5	Revision	1	Anterior and central	Glenoid wear status post prior hemiarthroplasty	Yes	25.3	Radial nerve palsy, underwent tendon transfers at 12 months postoperatively
5	74	M	Right	24.8	Primary	0	Central and posterior	Rotator cuff tear arthropathy	Yes	16.4
6	63	M	Left	31.7	Primary	0	Central and posterior	Rotator cuff tear arthropathy	Yes	15.1
7	60	M	Right	35.8	Primary	0	Global	Neuropathic arthropathy	Yes	17.3	Humeral component loosening at 4 months, underwent revision of humeral component alone at 4 months postoperatively
8	69	F	Left	20	Revision	1	Global	Glenoid wear status post prior hemiarthroplasty	Yes	15.9
9	44	F	Left	30.6	Revision	2	Global	2-stage revision of prior infected rTSA	Yes	20.3
10	83	F	Left	34.6	Primary	0	Central and posterior	Rotator cuff tear arthropathy	Yes	22.9
11	58	F	Right	23.7	Revision	3	Global	2-stage revision of prior infected rTSA	Yes	56.8
12	73	M	Right	30.8	Revision	1	Central and posterior	Glenoid wear status post prior hemiarthroplasty	Yes	38.1	Deceased at 40 months postoperatively
13	49	F	Right	24.2	Revision	3	Global	2-stage revision of prior infected rTSA	Yes	6.1
14	63	F	Right	21.3	Revision	2	Central and posterior	Glenoid wear status post prior hemiarthroplasty	Yes	50.7

M: male; F: female; BMI: body mass index; TSA: total shoulder arthroplasty; rTSA: reverse total shoulder arthroplasty; ACDF: anterior cervical discectomy and fusion.

Figure 2.

Preoperative coronal (a) and axial (b) computed tomography (CT) slices of the same right shoulder, redemonstrating the end-stage shoulder arthritis and severe glenoid retroversion.

Eligibility for inclusion in this study required a minimum of 6-month clinical or radiographic follow-up. All included patients provided documented verbal or written consent for participation in this study. Demographic and comorbidity information for each identified patient, as well as their respective surgical history, was collected through retrospective chart review (Figure 3).

Figure 3.

2-Year postoperative Grashey (a), anteroposterior (b), and axillary (c) radiographs of the right shoulder in the same patient following their primary reverse total shoulder arthroplasty using a patient-specific glenoid baseplate.

Clinical and radiographic assessment

Clinical assessment of included patients involved pre- and post-operative assessments of range of motion (forward elevation (FE), abduction (AB), external rotation (ER), and internal rotation (IR)), Visual Analog Pain Scale (VAS) score, Single Assessment Numeric Evaluation (SANE) score, American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form (ASES) score, and Penn Shoulder score. Range of motion data and patient-reported outcome measures were collected through retrospective chart review when present. Telephone and/or electronic mail questionnaires were used to collect additional data when necessary.

Radiographic assessment was performed by one of two attending orthopedic shoulder arthroplasty surgeons and an orthopedic surgery resident to assess for signs of implant loosening, failure, or migration on the included patient's most recent three-view shoulder radiographic series in comparison to serial prior imaging studies and in conjunction with clinical evaluation.

Occurrence and characterization of complications and any re-operations were assessed by retrospective review of medical records and patient interviews. Re-operation was defined as having undergone surgical intervention on the rTSA components for any reason following surgery. Implant survivorship was defined as the lack of revision surgery involving explanation of the custom glenoid baseplate as confirmed by chart review or individual patient confirmation.

Preoperative planning and implant design

“VRS design is performed in close collaboration between the attending surgeon and implant design engineers based upon the patient-specific osseous anatomy. Both surgeons involved in this study used similar implant design parameters, and aimed for between neutral and 5 degrees of inferior tilt, 10 degrees of retroversion, and sought to restore glenoid baseplate lateralization to the remaining paleoglenoid. Both surgeons also accepted a minimum central screw length of 25 mm. The implant design in all cases provided for a minimum of three peripheral screws, but preferably allowed for four peripheral screws when feasible. This series includes both anterior overhang designs and boss designs, and it is notable that there are numerous instances in this series where the implant engineers mandated a boss design. It is one surgeon's preference to add a boss to the design only when necessary, and otherwise utilize an anterior flange to assist in implant stability immediately upon implantation. The other surgeon prefers to add a boss to the design whenever feasible.”

Surgical technique

All patients were placed in beach chair position with the operative extremity secured in a surgeon-controlled arm positioner. A deltopectoral approach was utilized in all cases. In the primary setting, a tenodesis of the long head of the biceps to the upper border of the pectoralis major tendon near its insertion on the humerus was performed. During the surgical approach, the subscapularis tendon peeled from its insertion on the lesser tuberosity and tagged for later repair, if possible, at the end of the case. In primary cases, the humeral head was cut using standard cutting guides. In revision cases, the prior humeral implants were removed and the intramedullary canal was cleared of cement, if present.

In all cases, adequate glenoid exposure was obtained, with soft tissue release performed at the discretion of the attending surgeon in order to allow the patient-specific implant model to sit completely flush on bone. In the instance that incongruity was present between the patient's glenoid and the model, a burr was used to contour the bone to allow satisfactory placement of the final custom implant.

Glenoid reaming then was performed in standard fashion using a cannulated system. In cases where a boss was indicated, the central boss reaming guide was used. The custom glenoid baseplate was then placed within the glenoid vault using the anterior lip to assist with orientation and seated completely. The glenoid baseplate was then secured using a central 6.5-mm non-locking screw and peripheral 4.75-mm locking screws. Excess bone about the baseplate was then excised to prevent the occurrence of bony impingement.

After intraoperative trialing and assessment of glenosphere size and offset, the final glenosphere was impacted into the custom baseplate. Humeral components were then placed in standard fashion after similar trialing. Aside from the patient-specific custom glenoid baseplate, all glenospheres and humeral components were standard implants from the comprehensive reverse shoulder arthroplasty system by Zimmer-Biomet (Warsaw, IN, USA).

Statistical analysis

Statistical analysis was performed using RStudio (RStudio Inc). Descriptive statistical measures (absolute counts, means, ranges, percentages) were used in reporting clinical outcomes. Differences between mean pre- and post-operative range of motion measurements were assessed. A one-tailed t-test was conducted to determine if there was any statistically significant difference in range of motion postoperatively. Significance was set at p ≤ 0.05.

Results

Patient population

Fifteen shoulders were identified as having undergone primary or revision rTSA using the VRS at our institution from 2017 to 2022. One patient was lost to follow-up after 1 month postoperatively and was excluded from final analysis. Therefore, 14 shoulders in patients with a mean age of 65.1 years were available for assessment of implant survivorship at mean final follow-up time of 26.6 (range: 6.1–56.8) months (Table 1). Five patients underwent primary rTSA using the glenoid VRS, whereas nine underwent revision surgery for a variety of indications, respectively, listed in Table 1. In patients undergoing revision rTSA, six patients had undergone one prior shoulder arthroplasty, two patients had undergone two prior shoulder arthroplasties, and one patient had undergone three prior shoulder arthroplasties.

Clinical outcomes

Range of motion assessments were available in 14 of 14 (100%) shoulders at mean 11.5 (range: 6.1–26.0) months (Table 2). All individual patients experienced improvements in shoulder forward flexion, abduction, and ER postoperatively. Among the included shoulders, mean range of motion improved significantly in FE from 62 to 106 degrees (95% CI: 81–130 degrees; p = 0.007), abduction from 41 to 100 degrees (95% CI: 75–125 degrees; p = 0.001), and ER from 11 to 36 degrees (95% CI: 23–49 degrees; p = 0.002).

Table 2.

An overview of preoperative and postoperative range of motion for each included patient within the current study.

Patients	Forward flexion	Abduction	External rotation	Forward flexion	Abduction	External rotation
	Preoperative range of motion (degrees)			Postoperative range of motion (degrees)
1	60	50	40	80	85	60
2	70	70	0	130	80	50
3	80	0	30	100	100	15
4	60	45	20	95	95	10
5	130	0	10	150	150	50
6	130	130	30	180	180	90
7	60	60	0	90	90	45
8	0	0	0	5	5	5
9	80	80	0	160	160	20
10	20	20	10	90	90	25
11	45	30	0	100	60	25
12	90	45	20	170	170	70
13	0	0	0	80	80	20
14	40	40	0	50	50	15
MEAN	61.8	40.7	11.4	105.7	99.6	35.7

Forward elevation (FE): p = 0.007; Abduction (Abd): p = 0.001; External rotation (ER): p = 0.002.

Patient-reported outcome scores were available in 7 of 14 (50%) shoulders included for final analysis at mean 21.8 (range: 15.1–29.6) months). In available patients, final VAS score was 1.29 (SD = 0.70), SANE score was 72.14 (SD = 28.64), ASES score was 77.14 (SD = 15.83), and Penn Shoulder score was 72.26 (SD = 17.23).

Radiographic outcome

Radiographic outcomes were available in 14 of 14 (100%) shoulders at mean 11.5 (6.1–26.0) months. At final radiographic follow-up, no patients demonstrated evidence of glenoid baseplate loosening, instrumentation failure, or migration of components.

Complications

Two patients experienced complications after surgery related to their rTSA. One patient had loosening of their uncemented humeral component with gross rotation of the implant radiographically. They underwent revision of their humeral component alone at 4 months postoperatively. Microbial cultures taken at the time of the humeral revision surgery yielded no growth. Furthermore, there have been no further complications or signs of infection at 14 months postoperatively after revision of the humeral component. Additionally, their VRS baseplate has been maintained without complication since the time of its implantation. One patient experienced a radial nerve palsy postoperatively, and underwent tendon transfers of pronator teres to extensor carpi radialis brevis, flexor carpi radialis to extensor digitorum communis, and palmaris longus to extensor pollicis longus at 1 year postoperatively to improve wrist and finger extension. At 1 year following these procedures, the patient achieved 10 degrees of active wrist extension and metacarpophalangeal extension of all digits beyond neutral. Of note, one other patient deceased at 40 months postoperatively for an etiology unrelated to their rTSA.

Discussion

The management of severe glenoid deficiency remains a challenge for surgeons in order to achieve appropriate implant positioning and fixation, thereby minimizing the risk of early mechanical failure.^1,3–5 Patient-specific, 3D-printed, custom glenoid components have been used with promising early results in patients in whom other techniques of managing severe glenoid deficiency may be inadequate.^{4,12,13,16–19,21,22} The most important finding in this study was that the use of the VRS resulted in good clinical and radiographic outcomes in our series of patients undergoing rTSA with severe glenoid deficiency, with no failures of the custom glenoid baseplate observed at early follow-up.

Other studies evaluating implant survivorship and risk of complication with the use of the VRS have been similarly encouraging. Bodendorfer et al. reported clinical and radiographic outcomes at minimum 2-year follow-up (mean 30 months) of 12 shoulders in 11 patients with severe glenoid deficiency who underwent primary (n = 7) and revision (n = 5) rTSA with the VRS. ¹⁶ At final follow-up, the authors reported that no patients experienced any postoperative complications or required revision surgery, and that all implants appeared clinically radiographically stable. ¹⁶ Rangarajan et al. also reported outcomes at minimum 1-year follow-up (mean 18.2 months) of 19 patients who underwent primary (n = 9) and revision (n = 10) rTSA using the VRS. ⁴ In their series, Rangarajan et al. reported four occurrences of complications, with only one patient requiring explantation of the glenosphere and VRS. Three other re-operations were performed (two humeral component revisions, one hematoma evacuation), but no other patients demonstrated evidence of clinical failure of the VRS or radiographic implant loosening. ⁴ Dines et al. also reported a lack of clinical or radiographic failure in two patients undergoing rTSA with the VRS at 4 years and 18 months postoperatively, respectively. ¹² This current study adds to a growing body of case series’ that these implants avoid early catastrophic failure and show stable radiographic appearance and favorable clinical outcomes at early follow-up. Longer-term survivorship of the VRS implants remains unclear at this time and larger, prospective series with defined preoperative inclusion criteria are still needed, Finally, while not directly applicable, other reports in the literature involving the use of different systems of patient-specific, custom glenoid implants to address severe glenoid deficiency have similarly been without early clinical failure.^{17–19,21,22}

With regard to clinical outcome, our study found similar occurrence of improvement postoperatively to those in prior reports. Both Bodendorfer et al. and Rangarajan et al. described clinically and statistically significant improvements in range of motion and functional outcome scores between pre- and post-operative measures.^4,16 Using a different computer-aided design (CAD) and computer-aided manufacturing (CAM) total shoulder replacement (TSR) system (Stanmore Implants Worldwide, Elstree, UK), Uri et al. (n = 11) and Uri et al. (n = 21) reported significant improvements in pain and functional outcome scores at 35 months and 3 years’ mean follow-up.^21,22 Chammaa et al. also reported significant improvements in pain, functional outcome scores, and range of motion at 5 years postoperatively in 37 patients who received the CAD–CAM TSR for primary rTSA. ¹⁹ Notably, however, mean postoperative FE in this cohort was 64 degrees, which was attributed to the constrained design of the implant. ¹⁹ Finally, Debeer et al. reported satisfactory pain and functional outcome scores in a series of 10 patients undergoing rTSA using the glenius glenoid reconstruction system (GGRS; Materialise, Leuven, Belgium) for severe glenoid deficiency. ²⁰ Debeer et al. did not report preoperative functional outcome scores or range of motion data postoperatively. ²⁰ Collectively, these findings are suggestive that clinically and statistically significant improvements in functional outcomes may be seen postoperatively with the use of patient-specific, custom glenoid components for addressing severe glenoid deficiency. Similar to evaluation of implant survivorship, however, larger, prospective studies are needed to better characterize clinical improvements seen with these implant systems and longer-term follow-up is needed to determine the maintenance of the functional improvement that is seen in the early postoperative period.

Limitations

This study has several limitations. First, the mean follow-up period is relatively short for adequate characterization of implant survivorship. Despite this, our mean follow-up period is similar to, or longer, than others using the VRS or other patient-specific, custom glenoid implants in the current literature. Additionally, reports of survivorship of the VRS beyond mid-term follow-up are limited at this time based upon the VRS only receiving US Food and Drug Administration 510(k) clearance in 2016. ⁴

Similar to other studies on patient-specific, custom glenoid components, the population of patients in this study is relatively heterogenous with regard to indication for surgery. Further investigation should use a more defined patient population in order to better control for treatment effect. Our study is also limited in its retrospective nature, relatively small sample size, lack of complete clinical follow-up, and lack of a control group for comparison, though other reports in the literature examining clinical and radiographic outcomes with use of the VRS are similar in this manner. Finally, while this study does offer a comparison of pre- and post-operative range of motion demonstrating a significant and consistent clinical improvement across the patient population, this does not serve as a complete surrogate for assessing differences in functional outcome. Due to the retrospective nature of this study, preoperative functional outcome scores were not present for patients, preventing determination of the magnitude of functional improvement by comparison to the postoperative values.

Despite these limitations, our findings are of high clinical relevance given the clinical challenge that severe glenoid deficiency presents to surgeons and the relative paucity of literature present on the safety and implant survivorship with the use of the VRS and other patient-specific custom glenoid implants. With the projected increase in number of primary and revision rTSA's, the need for treatment strategies to manage severe glenoid deficiency will also become greater. ²⁴

Conclusion

Use of the VRS in this series of patients with severe glenoid bone deficiency undergoing rTSA resulted in encouraging clinical and radiographic outcomes, with no failures in implant survivorship seen at early follow-up. Larger prospective studies with longer-term follow-up should be undertaken in order to better determine the efficacy and longevity of this implant.